February 2020
Daniel Scott Smith
danielscottsmith@stanford.edu

*---------
* PURPOSE
*---------

This package reproduces analyses and output for the paper: 

Social scientization and the schooling state in United Kingdom parliamentary discourse, 1803–1909


*--------------------
SOFTWARE REQUIREMENTS
*--------------------

Stata 16-- note packages & other contingencies may need to be installed separately.
Python 3.6-- note packages & other contingencies may beed to be installed separately.
Jupyter Notebook


*---------------------
DIRECTORY ORGANIZATION
*---------------------

"00_data"
most raw data; for sources, see manuscript.

"01_scripts"
all scripts to produce analyses and visuals.

"02_visuals"
all raw data visuals (models, figures); empty until replication is implemented.


*----------------------
DIRECTIONS TO REPLICATE
*----------------------
1. Run all cells in 'cr_west_vdem10.ipynb". This creates the requisite "west_vdem10.dta" using the file "vdem10.csv" published by the V-Dem project. For convenience, a version of "west_vdem10.dta" is pre-included in this replication package in the "00_data" subdirectory. It contains variables on all states in the 19th century West.


2. Request pre-processed speech text data from author ("uk_terms.csv"), which has been omitted from this replication package due to size (~1000 MBs). Run all cells in 'cr_uk_coded.ipynb". This creates the requisite "uk_coded.csv", which contains the coded schooling and science terms (among others) for all speeches in the corpus. For convience, a version of "uk_coded.csv" is pre-included in ths replication package.


3. Open Stata and change current directory to the "*/00_replication/01_scripts" subdirectory of the replication package on local machine. Then, type "do ssh_analysis" in the stata command line. This calls the script "cr_uk_ivs.do", which creates all system-wide independent variables for the UK by importing and mergin the diverse data sources, and then uses these in the subsequent analyses. Last, it calls the script "ssh_figures.do", which produces all figures. All resulting raw tables and figures can be found in "*/02_visuals".